Algorithm selection for data communications in a parallel active messaging interface (‘PAMI’) of a parallel computer, the PAMI composed of data communications endpoints, each endpoint including specifications of a client, a context, and a task, endpoints coupled for data communications through the PAMI, including associating in the PAMI data communications algorithms and bit masks; receiving in an origin endpoint of the PAMI a collective instruction, the instruction specifying transmission of a data communications message from the origin endpoint to a target endpoint; constructing a bit mask for the received collective instruction; selecting, from among the associated algorithms and bit masks, a data communications algorithm in dependence upon the constructed bit mask; and executing the collective instruction, transmitting, according to the selected data communications algorithm from the origin endpoint to the target endpoint, the data communications message.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A parallel computer that selects an algorithm for data communications for a collective operation in a parallel active messaging interface (‘PAMI’) of the parallel computer, the parallel computer comprising a plurality of compute nodes that execute a parallel application, the PAMI comprising data communications endpoints, each endpoint comprising a specification of data communications parameters for a thread of execution on a compute node, including specifications of a client, a context, and a task, the compute nodes and the endpoints coupled for data communications through the PAMI and through data communications resources, the compute nodes comprising computer processors operatively coupled to computer memory having disposed within it computer program instructions that, when executed by the computer processors, cause the parallel computer to function by: associating in the PAMI data communications algorithms and bit masks so that each algorithm is associated with a separate bit mask, each bit in each mask representing the presence or absence of a characteristic of a collective instruction to be executed by use of the algorithm associated with that mask; initializing the PAMI; and partially preconstructing, upon initialing the PAMI, a bit mask for each type of collective instruction; receiving in an origin endpoint of the PAMI a collective instruction, the collective instruction specifying transmission of a data communications message from the origin endpoint to at least one target endpoint; constructing by the origin endpoint a bit mask for the received collective instruction, each bit in the mask representing a characteristic of the received collective instruction, wherein constructing a bit mask for the received collective instruction further comprises constructing the bit mask for the received collective instruction from one of the partially preconstructed bit masks; selecting by the origin endpoint, from the associated data communications algorithms in dependence upon the constructed bit mask, a data communications algorithm for use in executing the received collective instruction; and executing the received collective instruction by the origin endpoint, including transmitting, according to the selected data communications algorithm from the origin endpoint to the target endpoint, the data communications message.
A parallel computer system optimizes data communication within a parallel application using a Parallel Active Messaging Interface (PAMI). The system pre-associates data communication algorithms with specific bit masks, where each bit represents a characteristic of a collective instruction. Upon receiving a collective instruction at an origin endpoint, the system constructs a bit mask representing the instruction's characteristics. This bit mask is quickly derived from preconstructed masks initialized with the PAMI. The system then selects the appropriate data communication algorithm based on the constructed bit mask. Finally, the collective instruction is executed, transmitting the message to the target endpoint using the selected algorithm, improving overall performance. The PAMI is comprised of data communication endpoints, each specifying client, context and task parameters.
2. The parallel computer of claim 1 wherein selecting a data communications algorithm further comprises: iteratively bitwise comparing with the constructed bit mask the bit masks associated with data communications algorithms until a match is found; and taking as the selected data communications algorithm the data communications algorithm associated with the bit mask that matches the constructed bit mask.
Building on the parallel computer's algorithm selection, the system selects the data communication algorithm by iteratively comparing the constructed bit mask for the collective instruction with the pre-associated bit masks. This comparison is done bitwise until a matching bit mask is found. The data communication algorithm associated with the matched bit mask is then chosen for executing the collective instruction. This ensures the most appropriate algorithm is selected for each collective communication, improving efficiency.
3. The parallel computer of claim 1 further comprising computer program instructions that cause the parallel computer to function by: initializing a Message Passing Interface (‘MPI’) communicator; and partially preconstructing, upon initializing the MPI communicator, a bit mask for each type of collective instruction; wherein constructing a bit mask for the received collective instruction further comprises constructing the bit mask for the received collective instruction from one of the partially preconstructed bit masks.
The parallel computer system further improves data communication by initializing a Message Passing Interface (MPI) communicator. During this initialization, it also preconstructs bit masks for each type of collective instruction within the MPI environment. When a collective instruction is received, the bit mask representing the instruction is built upon the partially preconstructed MPI bit masks. This allows for quicker bit mask construction by utilizing pre-existing structures within the MPI framework.
4. The parallel computer of claim 1 wherein: each client comprises a collection of data communications resources dedicated to the exclusive use of an application-level data processing entity; each context comprises a subset of the collection of data processing resources of a client, context functions, and a work queue of data transfer instructions to be performed by use of the subset through the context functions operated by an assigned thread of execution; and each task represents a process of execution of the parallel application.
In the parallel computer system, the PAMI is structured around clients, contexts, and tasks. Each client represents a collection of data communication resources dedicated to an application. A context is a subset of client resources, containing context functions and a work queue for data transfer instructions executed by a thread. A task represents the execution of a process within the parallel application. This layered structure allows for efficient management and allocation of resources within the parallel environment.
5. The parallel computer of claim 1 wherein each context carries out, through post and advance functions, data communications for the parallel application on data communications resources in the exclusive possession of that context.
Within the parallel computer system, each context handles data communication for the parallel application using "post" and "advance" functions. These functions operate on data communication resources exclusively possessed by that specific context. This exclusive access prevents conflicts and allows each context to manage its communication independently.
6. The parallel computer of claim 1 wherein each context carries out data communications operations independently and in parallel with other contexts.
In the parallel computer, each context executes its data communication operations independently and in parallel with other contexts. This parallel operation maximizes the utilization of available resources and reduces communication bottlenecks, leading to improved application performance.
7. A computer program product for algorithm selection for data communications for a collective operation in a parallel active messaging interface (‘PAMI’) of a parallel computer, the parallel computer comprising a plurality of compute nodes that execute a parallel application, the PAMI comprising data communications endpoints, each endpoint comprising a specification of data communications parameters for a thread of execution on a compute node, including specifications of a client, a context, and a task, the compute nodes and the endpoints coupled for data communications through the PAMI and through data communications resources, the computer program product disposed upon a computer readable storage medium, the computer program product comprising computer program instructions that, when installed and executed, cause the parallel computer to function by: associating in the PAMI data communications algorithms and bit masks so that each algorithm is associated with a separate bit mask, each bit in each mask representing the presence or absence of a characteristic of a collective instruction to be executed by use of the algorithm associated with that mask; initializing the PAMI; and partially preconstructing, upon initialing the PAMI, a bit mask for each type of collective instruction; receiving in an origin endpoint of the PAMI a collective instruction, the collective instruction specifying transmission of a data communications message from the origin endpoint to at least one target endpoint; constructing by the origin endpoint a bit mask for the received collective instruction, each bit in the mask representing a characteristic of the received collective instruction, wherein constructing a bit mask for the received collective instruction further comprises constructing the bit mask for the received collective instruction from one of the partially preconstructed bit masks; selecting by the origin endpoint, from the associated data communications algorithms in dependence upon the constructed bit mask, a data communications algorithm for use in executing the received collective instruction; and executing the received collective instruction by the origin endpoint, including transmitting, according to the selected data communications algorithm from the origin endpoint to the target endpoint, the data communications message.
A computer program product for a parallel computer system optimizes data communication using a Parallel Active Messaging Interface (PAMI). The program pre-associates data communication algorithms with bit masks, each representing collective instruction characteristics. Upon receiving an instruction at an origin endpoint, a bit mask is constructed, efficiently derived from preconstructed masks initialized within PAMI. The program then selects the appropriate algorithm based on the mask and executes the instruction, transmitting messages using the selected algorithm. The PAMI is comprised of data communication endpoints, each specifying client, context and task parameters. The computer program product is stored on a computer-readable medium.
8. The computer program product of claim 7 wherein selecting a data communications algorithm further comprises: iteratively bitwise comparing with the constructed bit mask the bit masks associated with data communications algorithms until a match is found; and taking as the selected data communications algorithm the data communications algorithm associated with the bit mask that matches the constructed bit mask.
Building on the algorithm selection within the computer program product, the selection process involves iteratively comparing the constructed bit mask for a collective instruction with the pre-associated algorithm bit masks. This bitwise comparison continues until a matching mask is found. The data communication algorithm associated with that matched mask is then selected. This ensures the most efficient algorithm is chosen for each collective communication, maximizing overall system performance.
9. The computer program product of claim 7 further comprising computer program instructions that, when installed and executed, cause the parallel computer to function by: initializing a Message Passing Interface (‘MPI’) communicator; and partially preconstructing, upon initializing the MPI communicator, a bit mask for each type of collective instruction; and wherein constructing a bit mask for the received collective instruction further comprises constructing the bit mask for the received collective instruction from one of the partially preconstructed bit masks.
The computer program product further enhances communication by initializing a Message Passing Interface (MPI) communicator. This initialization includes preconstructing bit masks for each type of collective instruction within the MPI environment. Subsequently, when a collective instruction is received, the bit mask construction builds upon these preconstructed MPI bit masks. This approach streamlines bit mask generation by leveraging pre-existing structures within the MPI framework.
10. The computer program product of claim 7 wherein: each client comprises a collection of data communications resources dedicated to the exclusive use of an application-level data processing entity; each context comprises a subset of the collection of data processing resources of a client, context functions, and a work queue of data transfer instructions to be performed by use of the subset through the context functions operated by an assigned thread of execution; and each task represents a process of execution of the parallel application.
Within the computer program product, the PAMI is structured with clients, contexts, and tasks. Each client represents a collection of data communication resources allocated to an application. A context is a subset of those client resources, containing functions and a work queue for data transfer instructions executed by a thread. A task represents the execution of a process within the parallel application, providing efficient resource management.
11. The computer program product of claim 7 wherein each context carries out, through post and advance functions, data communications for the parallel application on data communications resources in the exclusive possession of that context.
In the computer program product, each context manages data communication through "post" and "advance" functions, operating on resources exclusively possessed by that context. This exclusive access ensures that each context can independently manage its communications without interfering with other contexts within the parallel system.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
December 3, 2010
July 16, 2013
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.